Notice!

We moved!

Please visit us at our new location which can be accesed through http://forum.legendarena.com/ or via legendarena.mypunbb.com

Viewing Single Post From: Zoid's big list of suggestions
Zoid	Jul 1 2009, 11:13 PM
Coppersmith Posts: 78 Group: Members Member #234 Joined: Jul 1, 2009	Ok, here's the great big pile of suggestions I've been mulling over for the duration of the round. Some are more important than others (class balancing is a must), and some are easier than others. I'll try to designate which ones should be easy to code and which ones are more involved. Classes: It's pretty obvious the classes aren't at all balanced. If I had to measure them on a scale of 1-10, I'd say Rogues are a 10, Trademen are a 7, and Warriors are a 4 or 5. 32 of the top 50 levels are Rogues and 34 of the top 50 ranked players are Rogues. For those of you who think this is just because more people play Rogues, in ranks 51-250 there are 78 Rogues, or 39%. This is almost balanced (balanced would be 33%), but the top 50 players are 68% Rogues, so you cannot argue there are more top ranked Rogues because there are simply more of them. Even the area in which you'd expect tradesmen to dominate, current ranking (since this is based mostly on charm strength and level), Rogues still make up 60%. The reason for this is simple, Rogues start out more powerful and scale better than the other classes. Warriors don't scale very much at all (slightly, since they save more money the more life they have, but that's negligible) and Tradesmen scale roughly at a square root pace (since their bonus is basically ordering larger charms), while Rogues scale linearly (the finishing move does triple damage regardless of whether you're doing 1 damage or 100). Since the finishing move basically ensures that you'll win the close ones, a Tradesman or a Warrior must be considerably stronger to win regularly against a Rogue. This is more difficult to do than you'd think, since once everyone hits about level 60, even if you have 30 more total charm power (which is quite expensive, that's getting power 45 charms when Rogues get power 35, which is fairly reasonable to expect), you're still only maybe 10% more powerful, which isn't enough to make the difference. Further, in order to actually get 30 more charm power, you have to order charms 36 power stronger, to compensate for the +5 agl. Lastly, add in the random combat system and Rogues tend to be avoided and win more often on defense, giving them less experience loss. Even now, if I'm fighting a Rogue and a miss turns into a critical hit, it can easily change the battle from me winning with 250-350 life left to me losing. Simple Changes (easy to code, I think): Remove the agility bonus to Rogues. Leave it on the subclass without the finisher, for obvious reasons, but it's quite clear that they simply don't need it. Buff Warriors considerably. I'd say give them an extra +5 strength at level 10 and boost the healing to 1/2 (Warrior) and full (berserker). This also gives them a neat ability, that they can kill very quickly (since they don't need to heal), saves them money, and makes them unkillable if you can't kill them in one stamina. Trademen aren't too bad off, but still need a little bit more. They do get much better in the late game, for obvious reasons, so I would recommend boosting the inital skillpoints to 50 or 60 and not giving any for the subclass. This will help in the early game to even things out a bit more (enabling them to go for combat skill much sooner), but not too much in the late game when everyone has maxed skills anyway. More complex changes(not necessarily easy to code): Change the Rogue's finishing move to a first strike type move, to help with the random factor. At the start of a battle, they get a first strike for 2/3 damage (Rogue) or 1.5 times normal damage (operative), then the battle proceeds normally. The damage done is lower than the finishing moves, but they happen every time. Stats: Agility is simply too weak, which further exacerbates the Rogue issue (Rogues scale very well with strength). I would boost the effect of agility by about 15%. Another of the big imbalancing things about this round was that agility wasn't worth much and life simply didn't help against Rogues (you could boost life, but that just meant you'd be at 20% sooner). The Master Finishing Move can be treated as two extra attacks, which mean a lot less when every battle goes for 20 hits (so one person does 20 hits and the other 22) as opposed to 10 (so one person does 10 and the other 12). WP Score: I like this better than not having it, but I think it could use some tweaking. What would be fun is changing this to straight win % for individuals, but using it as another clan ranking. That way, a clan that suffers attrition (which a LOT do) halfway through the round could still have a shot at a top ranking slot. Obviously it would need some tweaking to make it non-exploitable, but it would be nice for a clan of only 8 people to be able to compete for a top slot. Personally, right now I feel as though WP score is too easy to exploit and too easy to mess up and knock yourself out of the rankings from the beginning (as I did, prayed 600 stamina into endurance at the start while in a clan, d'oh). Spamming: This used to be unpleasant when I played before, but now it's gotten downright ridiculous. We no longer have to F5 the market all day long to buy a decent charm, but it's no coincidence that the top 2 players were horrendous spammers. Yes, I know plenty of other games work so that the more you play, the better you are, but don't we want LA to be better than those games? You're not going to attract more players with "just another" game that in order to win you have to devote your life to it. Simple suggestion: Increase experience lost when killed by a lower level and decrease experience lost when killed by a higher level considerably, maybe as much as 50% each. The vast majority of the spamming is done by someone of high level to someone of not quite as high level (i.e. they find a soft target and just kill him until he dies of it). What this change will do is make it easier for those who have been spammed to get revenge (which is much more fun for everyone than simply banning spamming) and make it less profitable to spam one target unless you're very powerful. If you ARE very powerful, then it won't affect you very much since people will still have a hard time killing you, but if you just got to high level because you spammed the crap out of 1 or 2 players and aren't that strong, you'll get beaten back down very quickly. This change may also be accompanied by an increase in defensive experience gain (so that those who are high level because they're really strong won't be hurt much by it), but you have to be careful with that or people will use it to cheat. Complex suggestion: Again, I think most people who have been spammed would much rather get revenge on their tormenter than just make spamming illegal or undoable. Thus, I suggest that there be some threshold of number of kills per day that someone can kill a player (I would say max stamina/5) before they get penalized. The penalty would be that for a certain period of time, the spammer is worth 8 warpoints to EVERYONE, regardless of clan wars. As before, this makes it so that those who really are very powerful won't be hurt that much, but those who simply found a soft target and beat it senseless will get similarly beaten senseless. Sponsoring: This, plain and simple, needs to be against the rules. Individual ranks should be individual ranks, not how much your clan helped you ranks. I'm not sure about any coding that can be done for this, but rule changes are needed. I played another game for a while, Ogame (weird name, fun space empire builder) where there was a very simple but very good rule: You can transfer and help players as much as you want, as long as they're weaker than you are. Any transfer or help of a weaker player to a stronger one is against the rules, and it's pretty strictly enforced. As such, multies aren't of much use in there as even if you're not caught with the multi, you'll get banned for "pushing" anyway. A similar rule would be good here and would solve the "oh, but I just had 42 friends who don't play anymore who wanted to sell me their tokens for 5 credits each" problem. Top ranking: It's better than the old system of "whoever orders the biggest charms on the last day wins," but still lacking. I have two suggestions for making it better, both of about equal complexity I think: 1. Have the top ranking points scale with time elapsed. My suggestion would be that the top ranking points you get are normal*(1+days elapsed/30) basically increasing it by about 3% each day. This means at the end of month 1, you get twice as many points as at the beginning of the round. At the end of month 2, you get 3 times as many points. This properly puts the emphasis more on the end of the round without neglecting the beginning. 2. Keep top ranking the way it is, but divide "final" ranking into 3 parts, top ranking, level, and current ranking (remove level from top ranking calculation, of course). Top ranking would then be a measure of how strong you have been the entire round, current ranking is how strong you are now, and level is how well you did with that strength. Total ranking would then be all 3 summed with equal weight given to each. This would add a lot more interest to the end of the round and more calculation, such as do I weight gain 4000 to get higher in level or do I put those points into stats for higher current rank? Endurance: I think I've created a monster. Originally it was a new idea that had never been done before (in round 5). Now it's something that everyone has to do in order to compete. As such, I think it's against the primary description of the game now "choose your path..." A determined spammer can get a level with 60-70 stamina, so it feeds into itself much too well now. Suggestion: Decrease stamina gained and increase max stamina gained. I honestly would say give people 30-40 more max stamina but not any stamina at the moment. Reducing it back down to 5 stamina though would be sufficient to hamper this rat race at the end. I would advocate putting it at 5 stamina gained and 20 max stamina gained, so that people wanting to do it for the sake of not fighting often could do so but it wouldn't be hugely necessary to do it at the end of the round. Battle system: I think I suggested this before way back in round 5, but one of the major problems with random battles (which we all hate) is that the range of damage is so high. This is compounded by Rogues. You can get hit with 9 times the damage you did last time when fighting a rogue, which just makes things too random. Suggestion: Keep the average the same but lower the range to double, so damage of 3-9 would now be 4-8. Same average damage but a smaller range. It's still enough of a range so that some random factors would influence battles, but not as much as it is now. Alchemy: With the new alchemist skill, you can pretty much kill anyone as long as your charms aren't horrible. The problem is that if your targets come online and potion up as well, you're screwed. Suggestion: Remove the restriction that you can't use a potion if you have one running. Don't make them stack, just make it reset the duration at full strength. This was yet another thing making Rogues too powerful in the end game. Everyone had enough money to buy potions most of the time, and tradesmen had more than enough (I've been buying them almost all the time for the past 2 weeks and I have over a million credits to my name in the bank or in GC). Making it so you can potion up continuously will give tradesmen and warriors a much needed boost (warriors save a ton of money from healing in the end game) over rogues once the additional charm power simply isn't enough. Praying: This is also too random. It can make a huge difference in the early round if you pray a couple times and get something and then go fight, but you have to get lucky. In addition, praying for endurance is WAY too random, I got 30 stamina for my first two 150 stamina dumps and then 130 when I prayed 210 stamina. Suggestion: These should give a guarantee that is whatever their average is. So if you should, statistically speaking, get 2 endurance for each 5 you pray, then just make it that you get 2 endurance each time you pray. They can make or break and early round attempt at high level and rank, so they really need to be reliable. GC Auctions: I didn't like this at first, but it's actually pretty nice. For most of the round it doesn't matter, but it will still go dry toward the end and so GC becomes a more precious commodity. Vastly better than the old F5 the market all day long Suggestion: Works well as it is, but I would change it so that the highest unique bid wins instead of the lowest. This won't affect anything for most of the round, but when the market goes relatively dry it will make it so that money is now a factor, which makes it less of a chore to bid on auctions and gives an advantage to those who have planned well and managed their finances well. Wars: I've always thought that it's dumb that you PEACE a clan to hurt them. This also makes the final warpoint rankings fairly meaningless, since you aren't really in control of it (if nobody has a mutual with you, you aren't winning warpoints). Also, 9 wars is too many. I was in a 1 man clan and wasn't even in the top 20 levels for a while and when I declared war on UC and JL, they had the space to declare back. Who you war with should be a pretty strategic decision, and it's just not right now. Suggestions: If clan sizes stay the way they are, reduce wars to 5 again. Maybe even have them 5 regardless of how big the clans are. You may get fewer targets per war with smaller clans, but there are also fewer people competing for your targets. Also, make every war worth 8 warpoints. Perhaps give some sort of experience or other bonus to having a mutual, but the final clan rankings should be based on what you did, not whether or not nobody would war with you. End of Round: This has been annoying for as long as I can remember. It's like being in the last quarter of a basketball game and not knowing if you have 30 seconds left or 15 minutes. The game really needs to end at a set time. This round is a perfect example of this too, if the round had ended before about 18:00 (four hours after the earliest time yosh said it would), The Order would have won both warpoints and size. If it had ended at 21:00 UC would have won both. Since it went to the very end, I think UC won warpoints and TO won size. UC gambled that it would end later and it did, most of TO (including myself) gambled that it would end earlier and it didn't. The winner of the round shouldn't be the clan that makes the best guess to when yosh is going to get around to hitting the "off" switch. I think this is all I have right now. The game is really doing well and I don't want to sound too critical, I just want to see it better. Discussion and disagreement are welcome and I hope yosh takes some of these suggestions into consideration. I'd offer to help code but I think the game is written in a language I don't know (I know FORTRAN and C/C++ really well and java and assembly sort of well). Glad to see a great game back and at it again.


Zoid's big list of suggestions · Suggestions

Viewing Single Post From: Zoid's big list of suggestions

Zoid

Jul 1 2009, 11:13 PM

Coppersmith

Posts:: 78
Group:: Members
Member: #234
Joined:: Jul 1, 2009

Ok, here's the great big pile of suggestions I've been mulling over for the duration of the round. Some are more important than others (class balancing is a must), and some are easier than others. I'll try to designate which ones should be easy to code and which ones are more involved.

Classes:
It's pretty obvious the classes aren't at all balanced. If I had to measure them on a scale of 1-10, I'd say Rogues are a 10, Trademen are a 7, and Warriors are a 4 or 5. 32 of the top 50 levels are Rogues and 34 of the top 50 ranked players are Rogues. For those of you who think this is just because more people play Rogues, in ranks 51-250 there are 78 Rogues, or 39%. This is almost balanced (balanced would be 33%), but the top 50 players are 68% Rogues, so you cannot argue there are more top ranked Rogues because there are simply more of them. Even the area in which you'd expect tradesmen to dominate, current ranking (since this is based mostly on charm strength and level), Rogues still make up 60%. The reason for this is simple, Rogues start out more powerful and scale better than the other classes. Warriors don't scale very much at all (slightly, since they save more money the more life they have, but that's negligible) and Tradesmen scale roughly at a square root pace (since their bonus is basically ordering larger charms), while Rogues scale linearly (the finishing move does triple damage regardless of whether you're doing 1 damage or 100).

Since the finishing move basically ensures that you'll win the close ones, a Tradesman or a Warrior must be considerably stronger to win regularly against a Rogue. This is more difficult to do than you'd think, since once everyone hits about level 60, even if you have 30 more total charm power (which is quite expensive, that's getting power 45 charms when Rogues get power 35, which is fairly reasonable to expect), you're still only maybe 10% more powerful, which isn't enough to make the difference. Further, in order to actually get 30 more charm power, you have to order charms 36 power stronger, to compensate for the +5 agl.

Lastly, add in the random combat system and Rogues tend to be avoided and win more often on defense, giving them less experience loss. Even now, if I'm fighting a Rogue and a miss turns into a critical hit, it can easily change the battle from me winning with 250-350 life left to me losing.

Simple Changes (easy to code, I think):
Remove the agility bonus to Rogues. Leave it on the subclass without the finisher, for obvious reasons, but it's quite clear that they simply don't need it.

Buff Warriors considerably. I'd say give them an extra +5 strength at level 10 and boost the healing to 1/2 (Warrior) and full (berserker). This also gives them a neat ability, that they can kill very quickly (since they don't need to heal), saves them money, and makes them unkillable if you can't kill them in one stamina.

Trademen aren't too bad off, but still need a little bit more. They do get much better in the late game, for obvious reasons, so I would recommend boosting the inital skillpoints to 50 or 60 and not giving any for the subclass. This will help in the early game to even things out a bit more (enabling them to go for combat skill much sooner), but not too much in the late game when everyone has maxed skills anyway.

More complex changes(not necessarily easy to code):
Change the Rogue's finishing move to a first strike type move, to help with the random factor. At the start of a battle, they get a first strike for 2/3 damage (Rogue) or 1.5 times normal damage (operative), then the battle proceeds normally. The damage done is lower than the finishing moves, but they happen every time.

Stats:
Agility is simply too weak, which further exacerbates the Rogue issue (Rogues scale very well with strength). I would boost the effect of agility by about 15%. Another of the big imbalancing things about this round was that agility wasn't worth much and life simply didn't help against Rogues (you could boost life, but that just meant you'd be at 20% sooner). The Master Finishing Move can be treated as two extra attacks, which mean a lot less when every battle goes for 20 hits (so one person does 20 hits and the other 22) as opposed to 10 (so one person does 10 and the other 12).

WP Score:
I like this better than not having it, but I think it could use some tweaking. What would be fun is changing this to straight win % for individuals, but using it as another clan ranking. That way, a clan that suffers attrition (which a LOT do) halfway through the round could still have a shot at a top ranking slot. Obviously it would need some tweaking to make it non-exploitable, but it would be nice for a clan of only 8 people to be able to compete for a top slot.

Personally, right now I feel as though WP score is too easy to exploit and too easy to mess up and knock yourself out of the rankings from the beginning (as I did, prayed 600 stamina into endurance at the start while in a clan, d'oh).

Spamming:
This used to be unpleasant when I played before, but now it's gotten downright ridiculous. We no longer have to F5 the market all day long to buy a decent charm, but it's no coincidence that the top 2 players were horrendous spammers. Yes, I know plenty of other games work so that the more you play, the better you are, but don't we want LA to be better than those games? You're not going to attract more players with "just another" game that in order to win you have to devote your life to it.

Simple suggestion:
Increase experience lost when killed by a lower level and decrease experience lost when killed by a higher level considerably, maybe as much as 50% each. The vast majority of the spamming is done by someone of high level to someone of not quite as high level (i.e. they find a soft target and just kill him until he dies of it). What this change will do is make it easier for those who have been spammed to get revenge (which is much more fun for everyone than simply banning spamming) and make it less profitable to spam one target unless you're very powerful. If you ARE very powerful, then it won't affect you very much since people will still have a hard time killing you, but if you just got to high level because you spammed the crap out of 1 or 2 players and aren't that strong, you'll get beaten back down very quickly. This change may also be accompanied by an increase in defensive experience gain (so that those who are high level because they're really strong won't be hurt much by it), but you have to be careful with that or people will use it to cheat.

Complex suggestion:
Again, I think most people who have been spammed would much rather get revenge on their tormenter than just make spamming illegal or undoable. Thus, I suggest that there be some threshold of number of kills per day that someone can kill a player (I would say max stamina/5) before they get penalized. The penalty would be that for a certain period of time, the spammer is worth 8 warpoints to EVERYONE, regardless of clan wars. As before, this makes it so that those who really are very powerful won't be hurt that much, but those who simply found a soft target and beat it senseless will get similarly beaten senseless.

Sponsoring:
This, plain and simple, needs to be against the rules. Individual ranks should be individual ranks, not how much your clan helped you ranks. I'm not sure about any coding that can be done for this, but rule changes are needed.

I played another game for a while, Ogame (weird name, fun space empire builder) where there was a very simple but very good rule: You can transfer and help players as much as you want, as long as they're weaker than you are. Any transfer or help of a weaker player to a stronger one is against the rules, and it's pretty strictly enforced. As such, multies aren't of much use in there as even if you're not caught with the multi, you'll get banned for "pushing" anyway. A similar rule would be good here and would solve the "oh, but I just had 42 friends who don't play anymore who wanted to sell me their tokens for 5 credits each" problem.

Top ranking:
It's better than the old system of "whoever orders the biggest charms on the last day wins," but still lacking. I have two suggestions for making it better, both of about equal complexity I think:

1. Have the top ranking points scale with time elapsed. My suggestion would be that the top ranking points you get are normal*(1+days elapsed/30) basically increasing it by about 3% each day. This means at the end of month 1, you get twice as many points as at the beginning of the round. At the end of month 2, you get 3 times as many points. This properly puts the emphasis more on the end of the round without neglecting the beginning.

2. Keep top ranking the way it is, but divide "final" ranking into 3 parts, top ranking, level, and current ranking (remove level from top ranking calculation, of course). Top ranking would then be a measure of how strong you have been the entire round, current ranking is how strong you are now, and level is how well you did with that strength. Total ranking would then be all 3 summed with equal weight given to each. This would add a lot more interest to the end of the round and more calculation, such as do I weight gain 4000 to get higher in level or do I put those points into stats for higher current rank?

Endurance:
I think I've created a monster. Originally it was a new idea that had never been done before (in round 5). Now it's something that everyone has to do in order to compete. As such, I think it's against the primary description of the game now "choose your path..." A determined spammer can get a level with 60-70 stamina, so it feeds into itself much too well now.

Suggestion:
Decrease stamina gained and increase max stamina gained. I honestly would say give people 30-40 more max stamina but not any stamina at the moment. Reducing it back down to 5 stamina though would be sufficient to hamper this rat race at the end. I would advocate putting it at 5 stamina gained and 20 max stamina gained, so that people wanting to do it for the sake of not fighting often could do so but it wouldn't be hugely necessary to do it at the end of the round.

Battle system:
I think I suggested this before way back in round 5, but one of the major problems with random battles (which we all hate) is that the range of damage is so high. This is compounded by Rogues. You can get hit with 9 times the damage you did last time when fighting a rogue, which just makes things too random.

Suggestion:
Keep the average the same but lower the range to double, so damage of 3-9 would now be 4-8. Same average damage but a smaller range. It's still enough of a range so that some random factors would influence battles, but not as much as it is now.

Alchemy:
With the new alchemist skill, you can pretty much kill anyone as long as your charms aren't horrible. The problem is that if your targets come online and potion up as well, you're screwed.

Suggestion:
Remove the restriction that you can't use a potion if you have one running. Don't make them stack, just make it reset the duration at full strength. This was yet another thing making Rogues too powerful in the end game. Everyone had enough money to buy potions most of the time, and tradesmen had more than enough (I've been buying them almost all the time for the past 2 weeks and I have over a million credits to my name in the bank or in GC). Making it so you can potion up continuously will give tradesmen and warriors a much needed boost (warriors save a ton of money from healing in the end game) over rogues once the additional charm power simply isn't enough.

Praying:
This is also too random. It can make a huge difference in the early round if you pray a couple times and get something and then go fight, but you have to get lucky. In addition, praying for endurance is WAY too random, I got 30 stamina for my first two 150 stamina dumps and then 130 when I prayed 210 stamina.

Suggestion:
These should give a guarantee that is whatever their average is. So if you should, statistically speaking, get 2 endurance for each 5 you pray, then just make it that you get 2 endurance each time you pray. They can make or break and early round attempt at high level and rank, so they really need to be reliable.

GC Auctions:
I didn't like this at first, but it's actually pretty nice. For most of the round it doesn't matter, but it will still go dry toward the end and so GC becomes a more precious commodity. Vastly better than the old F5 the market all day long

Suggestion:
Works well as it is, but I would change it so that the highest unique bid wins instead of the lowest. This won't affect anything for most of the round, but when the market goes relatively dry it will make it so that money is now a factor, which makes it less of a chore to bid on auctions and gives an advantage to those who have planned well and managed their finances well.

Wars:
I've always thought that it's dumb that you PEACE a clan to hurt them. This also makes the final warpoint rankings fairly meaningless, since you aren't really in control of it (if nobody has a mutual with you, you aren't winning warpoints). Also, 9 wars is too many. I was in a 1 man clan and wasn't even in the top 20 levels for a while and when I declared war on UC and JL, they had the space to declare back. Who you war with should be a pretty strategic decision, and it's just not right now.

Suggestions:
If clan sizes stay the way they are, reduce wars to 5 again. Maybe even have them 5 regardless of how big the clans are. You may get fewer targets per war with smaller clans, but there are also fewer people competing for your targets.

Also, make every war worth 8 warpoints. Perhaps give some sort of experience or other bonus to having a mutual, but the final clan rankings should be based on what you did, not whether or not nobody would war with you.

End of Round:
This has been annoying for as long as I can remember. It's like being in the last quarter of a basketball game and not knowing if you have 30 seconds left or 15 minutes. The game really needs to end at a set time. This round is a perfect example of this too, if the round had ended before about 18:00 (four hours after the earliest time yosh said it would), The Order would have won both warpoints and size. If it had ended at 21:00 UC would have won both. Since it went to the very end, I think UC won warpoints and TO won size. UC gambled that it would end later and it did, most of TO (including myself) gambled that it would end earlier and it didn't. The winner of the round shouldn't be the clan that makes the best guess to when yosh is going to get around to hitting the "off" switch.

I think this is all I have right now. The game is really doing well and I don't want to sound too critical, I just want to see it better. Discussion and disagreement are welcome and I hope yosh takes some of these suggestions into consideration. I'd offer to help code but I think the game is written in a language I don't know (I know FORTRAN and C/C++ really well and java and assembly sort of well). Glad to see a great game back and at it again.

Zoid's big list of suggestions · Suggestions